Multidimensional scaling of consumer preferences for a public transportation system: An application of two approaches

Abstract

Consumer attitudes toward a proposed new public transportation system were assessed through the application of two multidimensional scaling models to data on preference choices for system attributes. Carroll's vector model and Kruskal and Carmone's nonmetric unfolding model were compared on theoretical and empirical levels to determine their utility for exposing the latent structure of attitudes for a public project. While the unfolding model was attractive because of a theoretical property, the vector model was able to uncover latent dimensions for the attitudes which could be related via discriminant analysis to socio-economic and demographic characteristics of the respondents. The vector model also produced an outcome which was more closely related to a unidimensional analysis of these data. Therefore, even though both the vector and unfolding models produced plausible geometric representations of the attitudes which arc expected to aid urban transportation planners in designing systems, the vector model produced the more acceptable outcome.